BRUEGEL

POLICY
CONTRIBUTION
ISSUE 2015/12
JULY 2015

REFORM MOMENTUM
AND ITS IMPACT ON
GREEK GROWTH
ALESSIO TERZI

Highlights

Telephone
+32 2 227 4210
info@bruegel.org
www.bruegel.org

Greece has an imperfect track-record of structural reform implementation. However, the poor growth outcome of the Greek programmes is also a consequence of
the timing and composition of reforms, which were not optimally geared towards
a speedy transition to a new growth model based on the private sector. While the
main responsibility for this lies with the Greek authorities, international institutions share the responsibility for the poor growth-enhancing effect of reforms.
In the current context, further structural reform efforts should be mainly targeted
at supporting Greece's speedy return to solid growth rates. This is not only because
poverty and unemployment have reached very high levels, but also for political
economy reasons: reforms must quickly be seen to be working in order to buttress
the consensus in favour of reform.
Further efforts should be made to improve Greeces business environment and to
liberalise product markets, in addition to shifting taxation away from labour and
towards consumption. Reforms to improve the quality of institutions should
continue and are very much needed in the Greek setting, while taking into account
that their demanding implementation might use up administrative capacity and
their impact on growth will only be seen over long time horizons.
Alessio Terzi (alessio.terzi@bruegel.org) is an Affiliate Fellow at Bruegel. An earlier
draft of this paper was prepared for the conference A new growth model for the Greek
economy, held in Athens on 3 June 2015. The author would like to thank Guntram
Wolff, Francesco Papadia and Zsolt Darvas for their comments.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

02

REFORM MOMENTUM AND ITS IMPACT ON

GREEK GROWTH
ALESSIO TERZI, JULY 2015
1 INTRODUCTION
Following the unexpected revelation of large shortfalls in its public accounts, Greece had no choice
but to apply for an International Monetary
Fund/European Union macroeconomic adjustment
programme in May 2010, because financial markets were no longer willing to refinance its debt.
In general, IMF-led adjustment programmes operate on the basis of three core principles: (i) securing external financing, (ii) adopting domestic
demand-restraining measures consistent with
available financing, and (iii) proceeding with structural reforms to promote growth and adjust in the
medium term (Mussa and Savastano, 1999). To
prevent moral hazard, lending is provided in
tranches, which are disbursed only following periodic reviews that confirm that the conditions
attached are being respected.
Given the multiplicity of objectives, it is hard to
take a clear-cut position on whether programmes
are successful or not (see Sapir et al, 2014). The
ultimate objective, however, is growth former
IMF managing director Michel Camdessus said in
the 1990s that it is towards growth that our programmes and their conditionality are aimed1.
From this standpoint, after five years of assistance, it is safe to conclude that the Greek programmes have not delivered anywhere close to a
satisfactory outcome. There are differences of
opinion, however, over the causes of this failure.
Some of the reasons most frequently given
include:
1. Our primary objective is
growth, Michael
Camdessus, former IMF
Managing Director, Statement before the United
Nations Economic and
Social Council in Geneva, 11
July 1990, cited in Przeworski and Vreeland
(2000).

1 The Original sin argument: Given the size of the

Greek economy and the imbalances it accumulated, the degree of macroeconomic adjustment needed was daunting. This translated into
debt projections (and hence a programme) that
the IMF staff even in 2010 considered largely
unsustainable (and hence unrealistic), only to

be over-ruled by the IMF Executive Board on

political grounds (Schadler, 2013). The first programme foresaw a GDP fall of 7.5 percent and
financing was calibrated to this scenario. When
the economy contracted sharply, Greece had
to undertake more fiscal consolidation,
because financing was provided in nominal
terms (see Darvas, 2012). This connects us to
points 2 and 3, below.
2 Austerity: Because of its public finance origin,
the Greek crisis was mainly tackled with fiscal
austerity measures, which weighed on growth
more than initially foreseen because fiscal multipliers were underestimated (Blanchard and
Leigh, 2013) and led to a deflationary spiral
(Mazzolini and Mody, 2014).
3 Delay in debt restructuring: The pretence that
the Greek public debt was sustainable, while it
was widely acknowledged not to be, led to a
highly uncertain situation in 2010-12. The
uncertainty meant that the private sector held
back from investing and caused capital outflows, undermining the financial sector and,
ultimately, the recovery.
4 Institutional gap: Connected to point 3 above;
Greece was the first euro-area country to apply
for a bailout, meaning that there was a high
degree of uncertainty in the early stages of the
Greek crisis. No crisis management or lending
facility was in place at European level and the
fear of a default and euro-exit was prevalent.
This again acted as a drag on investment and
ultimately growth (Pisani-Ferry et al, 2013).
5 Common currency trap: Greeces low productivity was so misaligned with prices that only a
significant devaluation could have helped to
restore its competitiveness and kick-start the
economy. And indeed when intervening in
countries with a fixed exchange rate, the IMF

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

03
more often than not advocates a de-anchoring2.
However, this was not possible within a monetary union and required a very sharp and protracted internal devaluation that will continue
to weigh on growth.
More recently, a sixth reason has been put forward: the degree of implementation of structural
reform. Aslund (2015), for example, claims that it
is not austerity that explains Greece's poor growth
performance, but rather the lack of progress on
reform implementation. This point was echoed by
Blanchard (2015). Terzi (2015), on the other
hand, uses World Bank and Organisation for Economic Cooperation and Development (OECD) indicators to show that significant steps were taken to
improve the countrys non-price competitiveness
during the programme years, a finding confirmed
by Darvas (2015).
The time is therefore ripe to analyse in fine detail
the conditions attached to the Greek programmes
and to look in particular at the degree of structural
reform implementation under the first and second
programmes, the speed at which implementation
took place, and the headings under which reforms
were enacted, especially compared to the other
euro-area programme countries: Portugal and Ireland3. Greece has implemented significant
reforms. However, both in terms of timing and
composition, they have not been optimally geared
towards a swift transition to a new growth model
based on the private sector.
1.1 Literature on the impact of structural
reforms
The literature on the growth impact of structural
reforms is very broad. However, selected key findings of relevance for the Greek case include:

least four to five years to materialise. This is not

the case for all reforms: tax reforms5, for example, are expected to have a significant positive
impact on output in the short run (Barkbu et al,
2012, IMF WEO, 2004). Similarly, restoring the
soundness of the financial sector is likely to
benefit growth already over a shorter-term horizon by lifting credit constraints.
2 Synergies exist between product and labour
market reforms. As shown by Bassanini and
Duval (2009), the impact of individual structural reforms will be greater the more marketfriendly the institutional environment. As such,
well-designed reform packages are likely to
yield stronger results than piecemeal reforms.
This result is corroborated by Anderson et al
(2013) who show (also see Figure 1) how the
effect of joint product and labour market
reforms is greater, particularly over a five-year
horizon, than the sum of the individual impacts.
This result is in line with ECB (2014)6.
3. Institutional reforms take time to implement
and are not a sine qua non for a swift return to
growth. The fact that a strong correlation exists
between high quality institutions and economic prosperity is unchallenged (see Appendix). However, building on the findings of
Hausmann et al (2005), Rodrik (2007) details
how growth accelerations have been possible
over the past decades with minimal institutional change. This is echoed by Haggard
Figure 1: GDP eect of selected structural
reforms in the euro-area periphery, percentage
point deviation from baseline
4.5
4

Tax reform
Labour market reforms
Product market reforms

3. Cyprus and Spain also

went through assistance
programmes, but are
excluded for the analysis
for different reasons. The
Spanish programme was
focussed on the banking
sector and did not involve
the IMF. As such, it is not
contained in the MONA
database. The Cypriot
programme is contained in
MONA but, is still at an early
stage of its implementation,
and is largely centred on
the banking sector (as
Ireland).
4. Usually modelled as a
reduction in final goods
price mark-ups.
5. Usually intended as a
shift of taxation away from
labour and towards consumption.

3.5

1 Product market reforms4 are those most likely

to have the greatest positive impact over the
short- to medium-run, particularly in the tradable sector. This is a finding corroborated by
various papers for industrialised countries (see
IMF WEO, 2004; Cacciatore et al, 2012), for the
euro area (Barkbu et al, 2012), for the euroarea periphery (see Anderson et al, 2013) and
for Greece (see Varga et al, 2013). However, as
Figure 1 shows, most of these effects take at

2. A similar argument was

made for Latvia, which however chose, on political
grounds, to keep the peg to
the euro and face a sharp
internal devaluation.

Product and labour reforms

2.5
2
1.5
1
0.5
0
Year 1

Source: Anderson et al, 2013.

Year 2

Year 5

6. Although most of the literature suggests there is a

clear complementarity
between product and labour
market reforms, it should be
mentioned that selected
theoretical papers, such as
Amable and Gatti (2004),
suggest a potential substitutability over the long run.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

04
(2000) who underlines how changes in organisational routines, the creation of new institutions (such as regulatory agencies or
tax-collection agencies), or fundamental
changes in existing bureaucratic organisations
take a long time to reach fruition even if initiated quickly. If administrative and reform
implementation capacity is constrained, and if
a quick return to growth is considered crucial,
priority should be given to other structural
reforms that are likely to be growth-enhancing
already in the short- and medium-term.
2 GREECES REFORM EFFORTS IN PERSPECTIVE
Quantifying structural reform efforts and consequentially determining the degree of implementation of IMF adjustment programmes are
notoriously problematic (Dreher, 2009). In line
with Ivanova et al (2003), this Policy Contribution
makes use of the IMFs Monitoring of Fund
Arrangements (MONA) database, which contains
detailed information on all measures taken as part
of programme conditionality and, in particular,
tracks whether structural reform targets were met,
partially met, met with a delay, waived, or not met,
by the time of each programme review. This allows
calculation of the percentage of structural conditionality that was implemented. This approach is
in line with work carried out on the topic by the
European Commission (see Deroose and Griesse,
2014) and the OECD (see OECD, 2012).

gramme countries. Figure 2 shows the composition of reforms under the IMFs structural benchmarks heading for the first and second Greek
programmes, Ireland and Portugal. At this stage no
account is taken of whether reforms were implemented or not.
Compared to Ireland and Portugal, both Greek programmes significantly emphasised the restructuring of the governments operations. This is
unsurprising given the public-sector origin of the
crisis in Greece. Public enterprise reforms were less
at the centre of the Greek programmes than in Portugal. The Irish programme was (unsurprisingly)
very focussed on financial sector restructuring.
There were also sharp differences between the
first and second Greek programmes. The former
was less focussed on the financial sector and
more on pensions, the civil service and public
enterprises than the latter. Moreover, other structural measures (which includes reforms aimed at
improving the business environment) gained
more importance in the second programme. This
is something already documented by Terzi and
Wolff (2014) using a term-frequency methodology to examine programme documents in order to
grasp the general direction of conditionality.
Figure 2: Conditionality by reform headings,
euro-area country programmes
100%
90%

Making use of MONA-based indicators of reform

implementation has clear pros and cons. On the
positive side, the granular nature of MONA allows
the types of reforms that were implemented at
each stage of the programme to be tracked in
detail. On the negative side, all reforms are
weighed equally, whereas some measures are
surely qualitatively and quantitatively more
important than others. To alleviate this concern,
we also cross-check the MONA-based findings
with alternative indicators of structural reform
timing and effort (Box 1 on page 9).

80%
70%
60%
50%
40%
30%
20%
10%
0%

Greece 1

Greece 2

General government

Before looking at reforms that have been implemented, it is interesting to look at how programme
conditionality differed in different euro-area pro-

Ireland

Central Bank

Civil service and public employment reforms, and wages

Pension and other social sector reforms
Public enterprise reform and pricing (non financial sector)
Financial sector

2.1 Structure of the programme

Portugal

Labour mkts, excl. public sector

Other structural measures

Source: Bruegel based on IMF MONA database. Note:

categories used in the chart (eg general government, central
bank) are the original headings used in the IMFs MONA
database. No filtering has been done for implemented or nonimplemented measures.

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

05
2.2 Implementation of structural reforms
Although programme design and tailoring to a
countrys economic situation is important, success will ultimately depend on whether structural
reforms are implemented. In order to assess the
degree of programme implementation, Figure 3
shows the breakdown of the MONA-based implementation index described in the introduction to
this section and in Ivanova et al (2003)7.
From this analysis, Ireland comes out as the clear
best performer with a close to perfect implementation record. Interestingly, however, Portugal and
Greece (taking the first and second programmes
together) do not differ much in the share of programme conditionality fully implemented, both
hovering around the 80 percent mark. But differences emerge in terms of non-implementation.
Portugal experienced a large share of delays but
implemented almost the full (non-waived) set of
conditions in the end. Greece did not implement
roughly 10 percent of the structural reforms discussed with the creditors under the first and
second programme, and delayed just as many.
It is worth noting a difference between the first and
second Greek programmes. The first saw a high
share of delays, but most of the reforms were ultimately implemented. This is in line with the expert
survey-based findings of Pisani-Ferry et al (2012).
The bulk of Greek non-implemented measures are
to be found in the second programme.
This indicates that the time dimension might be of
relevance. Figure 4 shows the pace of reform
implementation, displaying the number of new
reform measures implemented in Greece, Ireland
and Portugal between each review. Note that
because each programme had a different starting
date, each review had a different date, and Figure
4 shows the order of reviews for each country.
As can be seen, the pace of reform implementation started slowly in Greece and picked up after
the first review. This is a common trend in all programmes, because reforms require time to be
designed and implemented. Although systematically below the pace of implementation of Portugal
(except for the fourth review), Greece was broadly
in line with the other countries during the first pro-

gramme. However, implementation of reforms by

Greece lost momentum towards the end of the first
programme (5th review) and in the transition to
the second Greek programme, as noted by PisaniFerry et al (2012). As the new programme slowly
phased in, reforms under new headings had to
enter the legislative pipeline. As such, we observe
a slow down in the number of reforms implemented under the first review of the second Greek
programme. It must be noted that the pace of
reform implementation picked up steadily until
the last review, by when Greece was implementing more reforms (in absolute terms) than Portugal. However, by this point the reform gap,
measured as the area between the Portuguese
Figure 3: Breakdown of conditionality by
implementation record
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%

Greece 1

Greece 2

Greece, total

Implemented

Ireland

Portugal

Partially implemented

Implemented with delay

Not implemented

Source: Bruegel based on IMF MONA database.

Figure 4: Number of reforms implemented by

each review
40

Greece second programme

35
30

Portugal
25
20
15

Ireland

10
5

Greece first programme

0
0

10

11

Source: Bruegel based on IMF MONA database. Note: The xaxis uses ordinal numbers to underline that each review took
place at different points in time for each programme country.

7. Waived conditions are

not considered when
assessing the degree of
implementation.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

06
and Greek line, was wide: over the course of two
programmes and five years, Greece implemented
(also partially and with delays) a total of 166
reform measures. For Portugal, this number stood
at 189 over three years.

2.3 Timing and composition of reforms

Exploiting the granularity of the IMF MONA database to its fullest extent, Figures 6-8 detail the
timing and composition of reforms implemented8
in Greece under the first and second programmes,
and in Portugal, as a comparison. MONA identifies
whether reforms have been implemented and also
tracks them over time to verify that they are not
reversed later in the programme. This implies that
in Figures 6-8, the last review available will effectively offer an overview of the composition of all
reform measures implemented (and not reversed)
during the whole programme.

Figure 5 shows a timeline of concluded programme reviews for Greece, Ireland, and Portugal.
This is particularly important because another
quantitative indicator of implementation is
whether reviews are broadly equally phased (see
Dreher, 2009) ideally, reviews should take place
periodically (generally every quarter), with delays
indicating faltering programme implementation
that holds up the disbursement of subsequent
loan tranches. Once again, Ireland (and to a great
extent Portugal) offer two examples of programmes on track, with broadly equally phased
reviews. Greece experienced two big implementation delays. The first, identified also in Figure 5,
took place during 2012, when political instability
following the call for a referendum on the bailout
programme by Greek prime minister George
Papandreou led to two consecutive elections, in
May and June. The second setback took place
towards the end of 2013 and the first half of 2014
when, as acknowledged by IMF (2014), adjustment fatigue had kicked in and the government
majority in parliament was thinning. This also puts
into perspective the Greek implementation surge
identified in Figure 4: although Greece was indeed
implementing more reforms than Portugal by the
time of its last review, these were achieved over a
one year period while the Portuguese reviews
were repeated every two to three months. All in all,
it is fair to conclude that Greeces implementation
was far from perfect.
8. Throughout this section,
attention is only paid to
whether reforms were
implemented and when. No
distinction is made
between reform measures
that were implemented, or
implemented with delay.
9. The first Greek programme was replaced by a
second under the IMF
Extended Fund Facility in
March 2012. See
http://www.imf.org/external/n
p/sec/pr/2012/pr1285.htm.

In line with the findings of Pisani-Ferry et al

(2013), Figure 6 shows how at its inception the
Greek programme was almost entirely focussed
on restructuring the public sector: redesigning the
tax system and increasing transparency (under
general government), reducing the public wage
bill (under civil service reforms), preparing privatisation plans (under public enterprise reforms)
and containing pension spending. Reforms targeted at restoring confidence in the financial
sector started to take effect from September 2010
(R1) but were clearly not a main component of the
implementation effort, especially when compared
to Portugal (see discussion below). Labour market
reforms can be observed only from early 2011
onwards (R3).
As a consequence, and perhaps unsurprisingly,
almost two years into the Greek programme, in
March 2012, the memorandum attached to the
request for an extended arrangement9 included an
unambiguous recognition that what could be
observed was a good deal of primary fiscal
adjustment but only some improvements in unit

Figure 5: Timeline of programme reviews

Greece first programme
R0

R1

R2

R3

R0

Greece second programme

R4

R5

R0

R1R2

R3

R4 R5

R0

R1

R2

R1R2

R6

R7

R8

R3 R4

R9

R10

R5

R11

R12

Ireland
R3

R4

R5

R6

R7

R8R9 R10 R11

Portugal
Jan-10

Aug-10

Feb-11

Sep-11

Source: Bruegel based on IMF MONA database.

Apr-12

Oct-12

May-13

Nov-13

Jun-14

Dec-14

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

07
Figure 6: Composition of implemented reforms in
the rst Greek programme, breakdown by review
100%
90%
80%
70%

labour costs. The European Commission (in line

with the IMF) concluded that there was a need to
recalibrate the programme strategy towards a
growth-enhancing structural agenda, reducing the
focus on fiscal adjustment (European Commission, 2012).

60%
50%
40%
30%
20%
10%
0%

R0

R1

R2

R3

R4

R5

Figure 7: Composition of implemented reforms in

the 2nd Greek programme, breakdown by review
100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
R0

R1R2

R3

R4

R5

Figure 8: Composition of implemented reforms in

the Portuguese programme, breakdown by review
100%
90%
80%
70%
60%
50%
40%

By comparing Figures 6 and 7, we see a change in

the headline composition of structural reforms
implemented under the second programme for
Greece. The share of general government-related
reforms dropped to less than 30 percent of the
overall implementation envelope in R0 (only to
return to high levels at later reviews).
After being a big omission from the first programme, measures falling under a new category
of other structural reforms (under which reforms
improving the business environment and promoting competition are grouped10) started to be implemented from the start of the second programme,
although to a limited extent. Within this category,
efforts focussed on improving the business environment rather than liberalising product and service markets (Box 1). This is in line with the findings
of IMF (2014) which noted in its last Greek review
(R5) that progress on product and service market
reforms had until then lagged behind.
Labour market reforms, on which rather slow and
limited progress was made under the first programme, were a more important component of the
second programme from the outset. Finally,
restructuring the financial sector was straight
away a more significant component compared to
the first programme, and stayed significant
throughout.

30%
20%
10%
0%

R0

R1

R2

R3

R4

R5

R6

R7 R8R9 R10 R11

Key for Figures 6-8:

General government

Central Bank

Civil service and public employment reforms, and wages

Pension and other social sector reforms
Public enterprise reform and pricing (non financial sector)
Financial sector

Labour mkts, ex. public sector

Other structural measures

Source for Figures 6-8: Bruegel based on IMF MONA database.

Note: R1 = first review, R2 = second review, etc.

Figure 8 shows the composition of reforms implemented in Portugal, to provide a comparison with
a country in the euro area that, like Greece, had to
restore its competitiveness while dealing with
unstable public finances. Apart from the first
review, the Portuguese programme was characterised by less focus on reforming the governments operations compared to the Greek
programmes. For Portugal, restructuring the financial sector took a central role from the inception of
the programme and remained an ongoing priority.
Reforms aimed at improving the labour market
began to be implemented very early on (R1) in the

10. Examples of reforms

falling under this heading
are liberalising closed professions, liberalising product markets, removing
barriers to competition in
the tourism/retail sector.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

08
programme. Reforms strictly aimed at restructuring the pension system11 and civil service12 kicked
in only at a later stage in the programme.
Finally, compared to both the first and the second
Greek programmes, a larger share of Portugals
implementation effort was devoted from the outset
to improving the business environment and promoting competition (under other structural measures). In Greece, product/business environment
reforms were delayed until late 2012.
2.4 Structural reforms and growth in Greece
The preceding sections have shown that neither
the timing nor the composition of reform implementation in Greece was optimally geared
towards a swift and strong return to growth. Under
the first Greek programme, the priority was given
to measures aimed at restructuring the governments operations, reorganising the civil service,
streamlining the budgeting procedures and
increasing fiscal transparency. These are all positive measures aimed at improving the quality and
effectiveness of decision-making but very
demanding in terms of implementation over the
relatively short horizon of a programme, and likely
to benefit growth only over the longer term.

11. For example,

introducing a legislative
proposal that would align
the rules and benefits of the
public sector pension fund
with the general pension
regime (MEFP 8). Done
(with delay) at R8R9.
12. For example passing a
public administration labour
law aimed at aligning the
public employment regime
with private sector rules,
including for working hours
and holiday time, and
termination of tenure (MEFP
8). Partially implemented
at R8R9 and then
implemented (with delay)
at R10.

This is particularly problematic in a country that,

already at the inception of the programme, displayed very low levels of administrative capacity,
bureaucratic quality, and hence implementation
capacity. To substantiate this claim, Figure 9
shows a World Bank expert-survey-based indicator aimed at capturing a countrys quality of civil
Figure 9: Government eectiveness indicator, 2.5 (min) to 2.5 (max)
2.5
2
1.5
1

service, degree of independence from political

pressures, quality of policy formulation and implementation. Two interesting remarks: first, Greeces
administrative capacity was much lower than all
the other programme countries: less than half of
Portugals before the crisis. Second, notwithstanding all the effort put into improving institutional quality, Greece has suffered a deterioration
(if at all a change) in its administrative capacity.
As explained above, institutions are hard to
change and improve over the relatively short duration of a programme. This would suggest that when
dealing with limited administrative capacity, it is
better to focus reform efforts on growth-enhancing measures, rather than first trying to improve
institutional capacity in the hope that a more wideranging economic programme can subsequently
be implemented.
Of course, some of these government-related
reforms, if not most, were considered necessary
not for growth-enhancing considerations, but for
fiscal consolidation purposes. However, as Rodrik
(2007) put it, institutional reforms, as well as a
consolidation of fiscal accounts, can be much
easier to undertake in an environment of growth
rather than stagnation or depression.
Reforms associated with a beneficial effect on
output only over long time horizons, such as institutional reforms, were a large part of the Greek
structural reform conditionality relative to other
euro-area programmes, such as Portugals. Tax
reforms, intended as a shift of tax burden from
labour to consumption, were not really observed.
VAT rates increased, but this was not coupled with
a reduction in the tax wedge, which remains one of
the highest in the OECD (IMF, 2014). Moreover,
measures aimed at improving the business environment and promoting competition were to a
great extent lacking, while labour market reforms
progressed independently, hence failing to exploit
the synergies between the two.

0.5
0

Cyprus

Ireland

Portugal

Greece

-0.5
-1
-1.5
-2
-2.5

Source: World Bank Governance Indicators.

2010
2013

Some of these shortcomings were reversed only

two years down the road, in the transition to the
second Greek programme, which emphasised
business environment reforms more, included
some continued effort on labour market reforms
and aimed to restore confidence in the financial
sector.

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

09
BOX 1: Alternative indicators of reform timing and effort
International organisations have assembled a wide variety of competitiveness indicators that, in some cases, can
be used to gauge structural reform efforts. However, like the MONA-based method employed in this paper, their use
has significant limitations.
First, only a few of these indicators are based on legislative changes and not cyclical elements that are beyond
the direct control of a government. For example, the World Economic Forum produces a Global Competitiveness
Indicator which, among its sub-components, includes a measure of product market efficiency. This might at first
sight look like a good indicator to measure effort and progress in product market reform. However, this pillar is
constituted by elements such as imports (as a % of GDP), which are likely to fluctuate due to a number of cyclical
factors, particularly during sharp recessions.
Second, assembling regulatory-based indicators of competitiveness is particularly burdensome and, as such,
these are often not produced on a yearly basis. For example, the OECDs indicator of Product Market Regulation
(PMR) is only published every five years (the last vintages being 2008 and 2013). This makes its use problematic in this setting, because it does not allow the timing of reforms to be gauged.
Third, these indicators are usually very specific and enable the zooming in on one specific characteristic of a
policy field. For example, the OECDs Employment Protection Legislation (EPL) index focusses on flexibility of
labour markets but is not well equipped to capture all reforms taking place under the labour-market heading. For
example, changes to the minimum wage legislation or active labour market policies would not alter its score. As
such, even a combined reading of these competitiveness indicators hardly gives a better overall assessment of
wide-ranging reform packages, such as those implemented by a country under an IMF/Troika programme.
With these caveats in mind, Table 1 below illustrates the change in score of the available yearly regulatory-based
competitiveness indicators as a proxy for reform progress13 in Greece and Portugal.

Table 1: Change in selected competitiveness indicators

2010

Greece
World Bank Doing Business index
Portugal
Greece
Employment Protection Legislation (EPL)
Portugal
Greece
Energy, Transport, and Communication
Regulation (ETCR)
Portugal

-2.0
2.4
-0.6
0.0
-0.06
-0.06

2011

0.5
1.0
0.0
-0.6
-0.02
0.00

2012

2.0
1.4
0.0
-0.3
0.00
-0.13

2013

2.4
-0.2

2014

1.7
0.0

total

4.6
4.6
-0.6
-0.9
-0.08
-0.19

Source: Bruegel based on World Bank, OECD.

The first indicator analysed is the World Banks Doing Business index, which captures the quality of the business
environment. This is a particularly good indicator for our purposes as it is a wide-ranging composite index, meaning that countries could experience similar improvements while crafting reform programmes that target countryspecific bottlenecks. Interestingly, Portugal and Greece experienced exactly the same improvement in their DB
index over the last five years, but with a difference: Portugal concentrated its reform efforts in the early years of
the crisis, whereas substantial improvements in Greece started materialising only from 2012 onwards (under the
Second Greek Programme). This chimes well with our MONA-based assessment that reforms aimed at improving
the business environment in Greece were delayed and, consequentially, their beneficial impact on growth was
delayed.
The OECDs EPL shows how both Portugal and Greece liberalised their labour markets to a comparable extent in the
first year of their respective programmes (in December 2010 in the case of Greece, which is why in MONA this is
recorded under the third review in February 2011). However, unlike in Greece, in Portugal a strong reform momentum was carried through to the second year of the programme. In 2012, the EPL index for Greece marks a small
improvement, which might seem to contrast with our MONA-based assessment. This is however because, rather
than liberalising its labour market, Greece was reforming its minimum wage legislation, hence slipping under the
EPLs radar. Unfortunately data for 2013 and 2014 is not available, because of the OECDs publication calendar
and by virtue of the way EPL-based reform effort is computed .
The third indicator shown in Table 1 is the OECDs Index of Regulation in energy, transport, and communication. We
see how after a comparable liberalisation effort in 2010, reforms under this heading broadly came to a halt in
Greece, whereas Portugal made significant progress in 2012. Once again, unfortunately, data limitations do not
allow us to capture the extent to which reform momentum sped up under the second Greek programme.

13. The associated years

are not those of publication
but rather, depending on
the methodology of construction of each indicator,
the ones in which reforms
took place. For example, the
OECDs EPL assesses the
state of labour market regulation on 1 January each
year. The change between
2012 and 2013 are hence
marked as reform effort in
2012.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

10
However, the timing and composition of reforms
has led to a sharp adjustment of labour costs but
less so of prices (Manasse, 2015). As Manasse
(2015) explains, a fall in nominal wages (which is
what matters for competitiveness purposes) that
was not matched by a reduction in prices also
implied a sharp contraction of real wages, which
aggravated the recession rather than promoting
growth. Synergies between labour and product
market reforms went largely unexploited in
Greece, with adverse effects on growth.
Portugal, on the other hand, adopted a timing and
composition of reforms more in line with the three
principles outlined in section 1.1. First, short- and
medium-term growth enhancing reforms started
already to be put in place from the first review (R1)
onward. Second, labour market and product
market reforms proceeded hand-in-hand (Figure
8 and Box 1). Third, reforms aimed at restructuring the operations of the government, including
improving the civil service, did not occupy a predominant role or were phased in at a later stage in
the programme. All in all, from a strictly growth perspective, this was a good recipe, which partially
explains why reforms in Portugal started producing earlier and better results than in Greece.

least at the beginning, too few measures were

taken to reorient Greece swiftly towards a new
growth model based on the private sector.
This shortcoming was to a certain extent
reversed in the second Greek programme,
which put somewhat more emphasis on product markets, competitiveness, and growth.
However, at this stage, amid a deep recession
and political instability, implementation
slipped.
In terms of the pace of reform, Greece started
at the same speed as other euro-area countries
but experienced significant delays in 2012 and
late 2013, and a significant reform gap (in
absolute terms) developed, in particular,
between Greece and Portugal.
By comparison, reforms in Portugal experienced some delays but there was ultimately a
better overall implementation rate, which was
more balanced between consolidation and
growth and, within the latter, between product
and labour market reforms. This implies that
reforms implemented in Portugal would more
quickly and strongly take effect than those in
Greece.

3 CONCLUSIONS
After five years of financial assistance to Greece,
exports are not yet picking up significantly, public
debt hovers above the 170 percent of GDP mark,
unemployment remains above 25 percent, and
GDP remains some 25 percent below its pre-crisis
level. Against this background, and compared to
the experiences of other countries going through
Troika adjustment programmes, few would define
the Greek programmes as successes. However,
opinions differ on the causes of the failure. In relation to one possible cause the relatively slow
implementation of reforms in Greece, and the
composition of those reforms this Policy Contribution has shown that:
Reform efforts under the first Greek programme
were very much focussed on restoring confidence in the countrys fiscal accounts. This was
to be achieved through both fiscal consolidation and institutional changes aimed at increasing fiscal transparency and accountability. At

Concluding that the Greek programme failed only

or mostly because reforms were not implemented
would be unfair to the Greek authorities. The analysis above suggests instead that implementation
was broadly on track in the early stages of the first
Greek programme. However, it was not focussed
enough on short- and medium-term growthenhancing reforms, as admitted by European
Commission (2012). This initial design failure was
then coupled with a limited administrative capacity, government instability and (at times) limited
ownership of the reform agenda by national
authorities. All these elements are associated in
the literature with faltering implementation of conditionality during adjustment programmes (Bird
and Rowlands, 2001; Dollar and Svensson, 2000;
Ivanova et al, 2003; Kahn and Sharma, 2001;
Nsouli, Atoian and Mourmouras, 2004) and placed
limits on the amount of reform that could credibly
be asked from Greece. One can conclude that
while the main responsibility lays with the Greek
national authorities, international institutions

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

11
have to a certain extent a shared responsibility for
the poor growth-enhancing impact of structural
reforms in Greece.
At the current juncture, the challenge for Greece
and its creditors will be to strike the appropriate
balance between the necessary fiscal consolidation, also dictated by the political or market constraints on the external financing that can be
granted to Greece, the desire to prevent blatant
injustice, and the need to return the country to
growth. Growth should be prioritised not only on
the ground that poverty has reached unacceptably high levels, but also on political economy
grounds. As explained by Williamson (1994),
there is a need to demonstrate at least one area in
which a reform programme has succeeded (and
an important one would be a return to strong
growth) in order to sustain the reform process. And
everyone seems to agree that Greece still needs
substantial reform.
In terms of composition of structural reforms to
achieve the goal of a speedy and solid return to
growth while waiting for past reforms to bear fruit,
some key takeaways are:
1 Product market liberalisation and efforts to
improve Greeces business environment will be
beneficial to growth, whereas labour markets
have been liberalised quite substantially and
are now broadly in line with the OECDs average,
as shown by Darvas (2015).
2 Institutional reforms should continue but bearing in mind that administrative (and hence
reform implementation) capacity is low and
constrained in the short-run. As such, short-run

growth enhancing measures should be given

priority.
3 Tax reforms, intended as a shift in the tax
burden away from labour and towards consumption, can boost output already in the short
run. Within this framework, an increase in the
VAT rate could very well be beneficial for growth,
but only if coupled with a reduction in the tax
wedge, which has not so far materialised in
Greece.
Clearly, an underlying crucial caveat of this analysis is its partial nature. It focusses only on growth
as the objective of an adjustment programme,
whereas programmes need to take account of
multiple realities, ranging from the fiscal situation,
to market sentiment and political feasibility in
both the creditor and programme countries, especially within a monetary union.
Moreover, within the growth target, this Policy
Contribution isolated and focussed only on
structural reforms. It is likely that all the factors
given in the introduction, from large fiscal
multipliers to heightened uncertainty, contributed
to the failure of the Greek programmes. To a certain
extent, these factors most likely mutually
negatively affected each other, for example with
austerity weighing on growth and social cohesion,
political stability, and ultimately structural reform
implementation. Isolating and analysing these
intricacies should be a matter of great interest for
future research, especially as, for good or for bad,
conditionality and financial assistance have been
ingrained in the European crisis management
framework and it cannot be excluded that they will
have to be activated again.

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

12
REFERENCES
Acemoglu, D., Johnson, S. and Robinson, J.A. (2001) The Colonial Origins of Comparative
Development: An Empirical Investigation, American Economic Review, 91(5): 1369-1401
Acemoglu, D. and Robinson, J.A. (2012) Why Nations Fail: The Origins of Power, Prosperity and
Poverty, New York: Crown
Amable, B. and Gatti, D. (2004) Labour and Product Market Reforms: A Case for Policy
Complementarity, IZA Discussion Paper, No. 1190
Aslund, A. (2015) Greeces Problem: Persistent Fiscal Irresponsibility and Too Few Reforms , PIIE
blog, 29 January
Barkbu, B., Rahman, J. and Valdes, R. (2012) Fostering Growth in Europe, Staff Discussion Note,
International Monetary Fund
Bird, G., and Rowlands, D. (2001) IMF lending: how is it affected by economic, political and
institutional factors?, Policy Reform, 4(3), 243270
Blanchard, O. (2015) Greece: Past Critiques and the Path Forward, iMFdirect, 9 July
Blanchard, O., and Leigh, D. (2013) Growth Forecast Errors and Fiscal Multipliers, Working Papers,
13/01, 1-43, International Monetary Fund
Bouis, R., Causa, O., Demmou, L., Duval, R. and Zdzienicka, A. (2012) The Short-Term Effects of
Structural Reforms, OECD Economics Department Working Papers, No. 49, OECD Publishing
Cacciatore, M., Duval, R. and Fiori, G. (2012) Short-Term Gain or Pain ? A DSGE Model-Based
Analysis of the Short-Term Effects of Structural Reforms in Labour and Product Markets, OECD
Economics Department Working Papers, No. 948, OECD Publishing
Darvas, Z. (2012) The Greek debt trap: an escape plan, Policy Contribution 2012/19, Bruegel
Darvas, Z. (2015) Is Greece destined to Grow?, Bruegel Blog, 16 June
Deroose, S. and Griesse, J. (2014) Implementing economic reforms - are EU Member States
responding to European Semester recommendations?, ECFIN Economic Brief 37
Dollar, D. and Svensson, J. (2000) What Explains The Success Or Failure Of Structural Adjustment
Programmes?, The Economic Journal, 110 (October), 894917
Dreher, A. (2009) IMF conditionality: theory and evidence, Public Choice, Vol. 141, 233-267
ECB (2015) Progress with structural reforms across the euro area and their possible impacts,
Economic Bulletin, Issue 2/2015, European Central Bank, Frankfurt
Helbling, T., Hakura, D. and Debrun, X. (2004) Fostering Structural Reforms in Industrial Countries,
World Economic Outlook, WEO, Chapter III, 103-146
IMF (2014) Fifth review under the extended arrangement under the extended fund facility,
Country Report No. 14/151, International Monetary Fund
Ivanova, A., Mayer, W., Mourmouras, A. and Anayiotos, G. (2003), What Determines the
Implementation of IMF-Supported Programs, Working Papers 03(8), 145, International Monetary
Fund
Kahn, M. S. and Sharma, S. (2001) IMF Conditionality and Country Ownership of Programs,
Working Papers 01(142), 129, International Monetary Fund
Manasse, P. (2015) Cosa andato storto e come rimediare, Il Sole 24 Ore, 11 June
Mazzolini, G. and Mody, A. (2014) Austerity Tales: the Netherlands and Italy, Bruegel Blog, 3
October
Mussa, M., and Savastano, M. (1999) The Approach to Economic Stabilization, NBER
Macroeconomics Annual, 14 (January), 79128
North, D. C. (1990) Institutions, Institutional Change and Economic Performance, New York,
Cambridge University Press.
Nsouli, S. M., Atoian, R. and Mourmouras, A. (2004) Institutions, Program Implementation, and
Macroeconomic Performance, Working Papers 04(184), 1, International Monetary Fund
Przeworski, A. and Vreeland, J.R. (2000) 'The effect of IMF programs on economic growth', Journal
of Development Economics, Vol. 62, 385-421

Alessio Terzi REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH

BRU EGE L

POLICY

CONTRIBUTION

13
Rodrik, D. (2009) One Economics, Many Recipes: Globalization, Institutions, and Economic Growth,
Princeton, NJ: Princeton Univerity Press
OECD (2012) Economic Policy Reforms 2012: Going for Growth, OECD Publishing
OECD (2013) Issues paper on corruption and economic growth, OECD Publishing
Pisani-Ferry, J., Sapir, A. and Wolff, G. (2013) EU-IMF Assistance to Euro-Zone Countries: An Early
Assessment, Blueprint 19, Bruegel
Sapir, A., Wolff, G., De Sousa, C. and Terzi, A. (2014) The Troika and financial assistance in the euro
area: successes and failures, ECON Committee Study, Brussels
Schadler, S. (2013) Unsustainable Debt and the Political Economy of Lending: Constraining the
IMFs role in Sovereign Debt Crises, CIGI Papers, No 19, October
Terzi, A. (2015) Can Greece become competitive overnight?, Bruegel Blog, 3 February
Terzi, A. and Wolff, G. (2014) A needle in a haystack: key terms in official Troika documents,
Bruegel Blog, 11 March

APPENDIX: THE RELATIONSHIP BETWEEN INSTITUTIONS AND GROWTH

One of the most established correlations in growth (or development) economics is that between quality
of institutions and wealth (North, 1990). As shown in Figure A1, richer countries are better governed.
The voluminous literature on the topic has gone to great length in trying to disentangle the two and
detect the direction of causality. Perhaps the most influential contribution to the field since North
(1990) is by Acemoglu, Johnson, and Robinson (2001) who, making use of an ingenious instrumental
variable approach, show how it is high quality institutions that cater for sustained growth and wealth.
However, as recognised by Acemoglu and Robinson (2012), whereas high quality institutions are
necessary in order to experience sustained periods of growth over long time frames, growth acceleration
episodes are possible also in poor institutional environments. This message is echoed by Rodrik (2007)
who argued that igniting growth is also possible with minimal institutional changes, whereas sustaining
it over long periods is more challenging and requires extensive institutional reforms.
To illustrate the point, Figure A2 on the next page shows the correlation between average post-global
financial crisis real GDP growth, and governance effectiveness one of the World Banks most widely
used indicators of institutional quality. As can be seen, the correlation coefficient is minimal, suggesting
that high growth is possible over relatively short time horizons (four years in this case) also in poor
institutional settings. In a way, this shows that poor institutions are not an unsurmountable brake to
economic activity, at least over a medium term horizon.

12

11.5

Figure A1: Correlation between quality of institutions and GDP per capita, world (LHS) and OECD (RHS)

NOR

ITA
CZE
GRC
SVK

10

Log GDP per capita

8
10

Log GDP per capita

10.5
11

LUX

HUN
POL

CHE
USA
NLD
AUT
IRL
SWE
DEU
AUS
DNK
BEL CAN
FIN
FRA ISL
GBR
JPN
ESP KOR
NZL
SVN
ISR
PRT
EST
CHL

TUR

9.5

MEX

-2

-1

0
1
Government effectiveness

Source: Bruegel based on World Bank. Note: Average over the period 2009-13.

.5

1
1.5
Government effectiveness

2.5

BRU EGE L

POLICY

CONTRIBUTION

REFORM MOMENTUM AND ITS IMPACT ON GREEK GROWTH Alessio Terzi

14
These findings are confirmed by more thorough quantitative studies, which also control for the initial
level of income, hence taking into account the possibility that what the correlation plots in Figure 3
capture is simple economic convergence. Once again, no statistically significant correlation between
institutional quality and (short- to medium-term) growth is observed (OECD, 2013). Similarly, more
case-based studies have tried at length to explain the East Asian paradox: high corruption levels did
not prevent countries such as China, Thailand, and Indonesia from experiencing strong output growth
during the 1980s through into the mid-1990s.
Extrapolating these findings to a macroeconomic adjustment programme setting suggests that if the
objective is to bring swiftly a country back to growth in order to improve its debt sustainability, reforms
overhauling the institutional setting are neither a necessary nor sufficient condition for success. Once
growth momentum is restored, however, improving the institutions will help to solidify and sustain it.

20

Figure A2: Correlation between quality of institutions and growth acceleration, world (LHS) and OECD
only (RHS)
TUR

CHL
EST

ISR
KOR

POL
SVK
JPN
GBR
FRA
HUN

CZE

IRL

AUS
CAN
USA
SWE
LUX
DEU
CHE
AUT NZL NOR
BEL
ISL

FIN

SVN
ESP
PRT

-10

-5

ITA

DNK

NLD

Average GDP growth

10
0

Average GDP growth

MEX

GRC

-2

-1

.5

Government effectiveness

Source: Bruegel based on World Bank. Note: Average over the period 2009-13.

1.5

Government effectiveness

2.5